Review of Machine Learning-Driven Studies of Tearing Modes in Tokamaks

A review article summarizing machine learning applications for predicting and controlling tearing modes in tokamak plasmas. The work was authored by Cristina Rea and Stuart Benjamin and last updated in May 2026. It focuses on advancements from large statistical data-analyses and ML algorithms aimed at enhancing plasma stability and performance.

Use Cases

• Training ML models for tearing onset prediction based on experimental tokamak data.
• Developing control algorithms to avoid and suppress tearing modes using statistical analyses.
• Interpreting the physics governing tearing onset through data-driven methods.
• Designing triggers for disruption mitigation systems like ITER's based on ML insights.

Strengths

• Authored by named researchers from the Plasma Science and Fusion Center.
• Systematically explains the challenge of tearing onset prediction and summarizes latest ML achievements.
• Focuses on advancements from large statistical data-analyses mentioned in the description.

Limitations

• The dataset is a review article; the underlying experimental data is not directly provided.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Plasma Science and Fusion Center Dataverse

Collection Method

Likely a compilation and analysis of published literature and experimental results.

Freshness

Last updated 2026-05-26 21:17:47; freshness should be verified.